Objective  To assess the effect of the use of containers with a sterilizing filter in autologous serum eye drops for the treatment of dry eye. Methods  Sixty dry eye patients were recruited for this prospective cohort study. Autologous serum was kept in containers with a sterilizing filter for 30 patients and， for the other 30 patients， autologous serum was kept in conventional containers. Subjects completed the Schirmer Ⅰ test （SⅠT， main outcome）， ocular surface disease index （OSDI）， fluorescein tear breakup time （BUT）， and corneal fluorescein staining （CFS） at baseline， and 2 weeks and 4 weeks after treatment. During home therapy， it is recommended not to use the same conventional container for more than 3 days， and the same containers with a sterilizing filter for more than 14 days. Sixty rabbits were selected for contamination test. Containers with a sterilizing fliter （30 rabbits） or conventional containers （30 rabbits） opened to give drops to rabbits. Contamination of autologous serum was tested at 3， 7， 14 and 21 days after use. A repeated measured ANOVA and Friedman repeated measured ANOVA on ranks were used for statistical analysis. Results  In the group using conventional containers， contamination was found in zero bottle at day 3， 2 bottles at day 7， 4 bottles at day 14 and 12 bottles at day 21 while no contamination was found in containers with a sterilizing filter. No significant difference was found in SⅠT， OSDI score， BUT and CFS between the two groups during 4 weeks treatment. Conclusion  Containers with an adapted filter can reduce the contamination rates in autologous serum eye drops and extend the use of autologous serum eye drops without affecting therapeutic efficacy.